summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2004-12-10pseudo-DMA for non-AV Quadras, inspired by NetBSDMartin Reindl
with lots of help from miod@ additional testing by nick@ and otto@ ok miod@
2004-12-09add missing line in history section that didn't make it through the firstMartin Reindl
syncing round pointed out by Dries Schellekens <gwyllion at ulyssis dot org>
2004-12-08spl1() -> splsoft(), and minor sync with other m68k trap.c; barely anyMiod Vallat
functional change.
2004-12-08Silence spurious interrupts unless option DEBUGMiod Vallat
2004-12-08Register serial interrupts on the right sources on AV machines, was brokenMiod Vallat
after the interrupt code change; from NetBSD.
2004-12-08Crank maxusers to 32.Miod Vallat
2004-12-08commented out {e,o}hci@cardbusMichael Shalayeff
2004-12-08add pcscp(4) at pciMartin Reindl
ok miod@
2004-12-08add ohci and ehci at cardbus to amd64, i386 and macppc. amd64 remainsDavid Gwynne
disabled for now. from deraadt@
2004-12-08Tweak interrupt handling code to allow shared interrupts for VIA2 sources.Miod Vallat
This will be necessary shortly.
2004-12-08Only map the DAFB video registers in attach, so that other code can mapMiod Vallat
the non-video parts. Will be necessary shortly.
2004-12-07remove old and unused NetBSD specific cruft.Brad Smith
2004-12-06set granularity for data segment used for struct cpu_info to bytes.Hans-Joerg Hoexer
ok mickey deraadt tedu niklas
2004-12-06emu(4) works; tybollt@solace.mh.seTheo de Raadt
2004-12-06use raw device; wbx@openbsd.deTheo de Raadt
2004-12-06Use uvm_grow() to account for stack growth, rather than home-grown codeMiod Vallat
or nothing. Inspired by a similar recent change in NetBSD.
2004-12-06Use config_defer().Brad Smith
From NetBSD ok miod@
2004-12-04add /etc/dhcpd.conf to FILES;Jason McIntyre
2004-12-03Always attach intvid even if the monitor sense code seems wrong; this isMiod Vallat
the case with non-Apple monitors connected with video adaptors.
2004-12-03Kill mac68k *SBC kernels. Instead, build kernels with both sbc scsi flavoursMiod Vallat
compiled in, but one disabled, and use config -e to swap them and build the other kernels; saves valuable compile time. Add ramdisks into the build (at release time). Since we are moving to a bsd.rd installation, it is not necessary to put kernels in .tgz files anymore, so don't bother, and adjust installation script sets information accordingly.
2004-12-02move IPL setup to intr.cMartin Reindl
ok miod@ >From: NetBSD
2004-12-02option DISABLE_EXT_CACHE is not used anymore.Miod Vallat
2004-12-02We do not support 2MB machines, so don't bother taking them into accountMiod Vallat
in allocsys().
2004-12-02Make BUFCACHEPERCENT adjustable through UKC, like all other arches do.Miod Vallat
(This means twice as less buffers on hp300 with unmodified GENERIC kernels now)
2004-12-02Oops, repair *SBC kernels.Miod Vallat
2004-12-02Add pci_decompose_tag() for sparc64.Brad Smith
From NetBSD ok deraadt@
2004-12-02enable vge(4) and ciphy(4); ok deraadtPeter Valchev
2004-12-01Much, much, much less stack pressure when reading a disklabel.Miod Vallat
This is a temporary workaround which might live longer than initially expected.
2004-12-01Do not store preciouss data below the kernel stack, or they risk becomingMiod Vallat
invisible and being lost...
2004-12-01No more splnone alias for spl0.Miod Vallat
2004-12-01Sync core dumping code with other m68k platforms.Miod Vallat
2004-12-01Let kernels with option DEBUG link again.Miod Vallat
2004-12-015 -> FC_SUPERDMiod Vallat
2004-12-01Remove unexpected (and wrong) easter egg.Miod Vallat
2004-12-01remove excessive chattiness from the bootblocks. our goal is to shrinkTheo de Raadt
them to < 5*1024, because the floppies are once again tight
2004-12-01Remove unused constants.Miod Vallat
2004-12-01VIA VT612x PCI Gigabit Ethernet adapter support, ok deraadtPeter Valchev
from FreeBSD
2004-11-30Move the struct wsscreen_descr from a per-driver global to a per-instanceMiod Vallat
field of the sunfb structure. This allows multiple instances of the same driver, but with different resolutions (such as a couple of vigra or a TGX cgsix and a TGX+ cgsix) to use distinct wsscreen_descr structures featuring different resolution information. Doing this allows more wsscreen_descr fiddling inside the sparc* fb api, and results in some code shrinkage (about 4KB on sparc GENERIC). [file excluded in previous commit by mistake. spotted by claudio@]
2004-11-30repair DEBUG kernelsMartin Reindl
ok miod@
2004-11-30provide a more accurate assembly delay() routineMartin Reindl
ok miod@ >From: NetBSD
2004-11-29Move the struct wsscreen_descr from a per-driver global to a per-instanceMiod Vallat
field of the sunfb structure. This allows multiple instances of the same driver, but with different resolutions (such as a couple of vigra or a TGX cgsix and a TGX+ cgsix) to use distinct wsscreen_descr structures featuring different resolution information. Doing this allows more wsscreen_descr fiddling inside the sparc* fb api, and results in some code shrinkage (about 4KB on sparc GENERIC).
2004-11-29plug memory leaks on errorPatrick Latifi
ok canacar markus millert
2004-11-29Port the remaining two sbus framebuffer drivers to sparc64; tvtwo untestedMiod Vallat
due to lack of hardware.
2004-11-29Spell precede correctly.Jonathan Gray
'looks fine' millert@, krw@. ok jmc@
2004-11-28Prevent ADB devices other than mices of generating "extended" protocol events.Matthieu Herrb
This fixes a problem with some iBooks were the touchpad generates a button6 event everytime the pad is touched. Based on my observations and a similar patch from Aymeric Vincent for NetBSD. Enhancements and ok miod@
2004-11-28Correct pmap_enter() flags usage.Miod Vallat
2004-11-28MAXSLP is not really an MD-configurable define so move it to param.h; miod@ ↵Michael Shalayeff
testing
2004-11-27use a pm_lock define instead of the long fluffMichael Shalayeff
2004-11-27Fix for building in another directory using config -s. ok deraadt@, miod@.Matthieu Herrb
2004-11-27In pmap_bootstrap(), replace PMAP_MD_RWZERO, which would leave the lowestMiod Vallat
page writeable, with PMAP_MD_RWLOW, which tells how many pages have to be left writeable on low addresses, since the mac rom needs more than one. This lets non-DDB mac68k kernels run. No change on non-mac68k platforms.